The major objectives of the project are to develop the TG.AC transgenic mouse model line (which carries a v-Ha-ras gene) as a model for (1) studying the multi-stage process of skin carcinogenesis: and (2) as a short -term (26 week) in vivo model for drug and chemical safety assessment and for prediction of the carcinogenic potential of chemicals. The induction of skin papillomas by repetitive topical application of a test chemical is considered to be the "reporter phenotype" that defines activity. But we and other laboratories have also demonstrated that other routes of administration, e.g., feed, water, oral gavage or I.P. injection are also appropriate for inducing tumors in the skin and other organm sites. Because of the chemical activity data base generated by LECM, the Tg.AC model has been accepted as a test component by the FDA for determining human risk assessment. The model detects both genotoxic and non- genotoxic carcinogens. Over 50 chemicals have been tested in the Tg.AC model to date. The results indicate that the model is most likely to detect trans-species carcinogens. All five of the human carcinogens (benzene, cyclosporin A, DES, melphalan, and TCDD) tested were detected. The model is most unlikely to detect non-genotoxic carcinogens that induce only liver or kidney tumors in a single sex/species, e.g., the female B6C3F1 mouse. Recent studies suggest that Tg.AC mice and the FVB/N parent strain are resistant to some classes of rodent liver carcinogens.Induced papilloma multiplicity (the reporter phenotype) has been shown to follow a reproducible dose-related pattern with such agents as benzene, TPA and TCDD. Thus Tg.AC mice are an appropriate model to evaluate the effects of diet, or the efficacy of anti-tumor agents or cocarcinogens that can modulate the tumor response. The exposure of Tg.AC mice to sodium arsenite in the drinking water significantly enhanced the multiplicity of TPA induced papaillomagenesis. The investigation of the role of early inflammatory or growth control response genes that associate with chemical or wound induced tumorigenesis in Tg.AC mice is continuing. The early induction of GM-CSF in the skin appears to be a requirement for TPA induced tumors. Injection of neutralizing antibodies to GM-CSF prior to TPA application significantly reduced tumor multiplicity, nearly to zero. Efforts are continuing to identify the key factors that activate the transgene in chemical and wound induced tumorigenesis.